CAVs as mobile safety actuators: authorization rules, performance evidence, and deployment pathways for managing oscillation-induced safety risks

Traffic oscillations generate localized and rapidly evolving rear-end conflict risks that are difficult to address through conventional infrastructure-centered traffic-management instruments. This study develops a selective connected and automated vehicle (CAV) authorization framework for managing oscillation-induced safety risks by treating qualified CAVs as temporary mobile safety actuators. The framework authorizes a small number of CAVs to perform bounded jam-absorption driving maneuvers only when observable oscillation states, vehicle positions, and safety constraints satisfy implementable operational rules. Traffic-flow theory and car-following dynamics are used to derive triggering conditions, absorbing-vehicle designation rules, no-secondary-jam constraints, termination conditions, and coordination logic for single-, two-, and multiple-oscillation scenarios. Scenario-based numerical experiments are conducted to evaluate safety-risk exposure, traffic smoothing and stability, traffic efficiency, mixed-traffic effects, and parameter sensitivity. Based on the intervention framework and scenario-based performance evidence, this study further identifies policy and practice pathways for translating vehicle-level automation capabilities into freeway safety-management decisions. Across representative scenarios, Time Exposed Time-to-Collision and modified Time Integrated Time-to-Collision decrease by 48.93–68.52% and 46.42–64.20%, respectively, while total travel time is reduced by 5.52–11.71%. Further analyses indicate that intervention effectiveness depends on CAV market rate, authorization delay, deceleration limits, and TTC-threshold settings, suggesting that these factors should be treated as policy and certification variables rather than purely algorithmic tuning parameters. The study also clarifies how CAV-based JAD can be embedded in traffic-management practice through event-based authorization, responsibility allocation, performance monitoring, and staged deployment. These findings provide policy-relevant evidence that selectively authorized CAV interventions can complement conventional traffic management by enabling localized, bounded, and auditable freeway safety governance.